Several factors affect the costs of both compliance and defiance and there-fore influence the outcome of a sanctions episode. The specific factors at play in each episode are described in the case histories. Several variables affect the costs of both defiance and compliance, albeit in opposite direc-tions. Here and in chapters 3 and 4 we divide the forces, somewhat artifi-cially, into two clusters: “political” and “economic” variables.
We focus on the following political variables:
䡲 Companion policies used by the sender country or coalition, namely, covert maneuvers (identified by a “J” in the case abstracts and in the tables in chapter 3), quasi-military activity (Q), and regular military activity (R).
䡲 The extent of international cooperation in imposing sanctions, scaled from 1 (no cooperation) to 4 (significant cooperation).
䡲 The involvement of an international organization, scored as 1 if both the sender and target countries are members of an international orga-nization that supports the sanctions and as 0 otherwise.
䡲 The presence of international assistance to the target country (indi-cated by an “A”).
䡲 The warmth of prior relations (before the sanctions episode) between sender and target countries, scaled from 1 (antagonistic) to 3 (cordial).
䡲 The political character of the target country’s government, scaled from 1 (autocracy) to 3 (democracy).
We focus on the following economic variables:
䡲 The cost imposed on the target country, expressed in absolute terms (US dollars), as a percentage of its gross national product (GNP), and in per capita terms.
䡲 Commercial relations between sender and target countries, measured by the flow of two-way merchandise trade between them, expressed as a percentage of the target country’s total two-way trade.
䡲 The relative economic size of the countries, measured by the ratio of the sender’s GNP to the target’s GNP.
6. Cited in the Oxford Dictionary of Quotation,3d ed. (Oxford: Oxford University Press, 1979, 152).
56 ECONOMIC SANCTIONS RECONSIDERED
䡲 The economic health and political stability of the target, measured by a judgmental scale from 1 (distressed country) to 3 (strong and stable country).
䡲 The type of sanctions used, namely, an interruption of exports from the sender country (identified by an “X”), an interruption of imports to the sender country (M), and an interruption of finance (F).
䡲 The cost to the sender country, measured by a judgmental scale from 1 (net gain to sender) to 4 (major loss to sender).
The variables just enumerated are described below in more detail. In chapter 3 we evaluate the connection between foreign policy success and the political variables. In chapter 4 we consider the relationship between success and the economic variables. In chapter 6 we summarize our find-ings and offer policy recommendations.
An important variable that we do not attempt to measure is the skill with which sanctions are employed as a foreign policy tool. For example, conflicting pressures within the sender country and its government can lead to an indecisive response, which neither emits the desired political signal nor imposes arduous costs on the target country. The classic exam-ple of confused signals was the League of Nations sanctions against Italy in 1935–36 (Case 35-1). The major powers in the League (the United King-dom and France) were torn between their desire to stop the Italian ad-vance in Abyssinia and their fear of upsetting the political balance in Eu-rope by driving Italy into an alliance with Germany. With an eye on upcoming national elections, British leaders in particular wanted to keep the peace in Europe; thus, while the League Council was considering sanctions, the decision was made to exclude oil, which would have had the most impact on Italy, and attempts were also made to appease Italy by ceding some territory in Abyssinia. In this case, as in many others, sanc-tions were largely intended to pacify domestic constituencies that were outraged over a foreign country’s behavior.
Political Variables
Companion Policy Measures
As Carl von Clausewitz (1832) famously observed, “War is nothing but the continuation of politics with the admixture of other means.”6 The same could be said of economic sanctions. Indeed, sanctions frequently serve as a junior weapon, or perhaps the starting gun, in a battery of
ANALYZING THE UTILITY OF SANCTIONS 57 diplomatic artillery aimed at the antagonistic state. Leaving aside the nor-mal means of diplomatic protest—recalling an ambassador or canceling a cultural mission—we distinguish three types of companion policies:
covert action, quasi-military action, and regular military action.
Covert action, mounted by intelligence forces, often accompanies the imposition of economic sanctions when the destabilization of a target government is sought. An example of unsuccessful covert action was CIA assistance in 1996 to Kurdish opponents of Saddam Hussein. In destabi-lization cases and episodes where major policy changes are sought, the sender state may also invoke quasi-military force: for example, massing troops at the border or stationing war vessels off the coast. Finally, sanc-tions may precede or accompany actual armed hostility, as happened in Panama, Haiti, Iraq, and Bosnia. In fact, political scientist David Baldwin (1985) conceives of sanctions as just one element in the “force curve” im-plemented to resolve conflicts between nations.
International Cooperation
In high-profile cases, such as the two world wars, the League of Nations’
foray against Italy, the series of US sanctions against the Soviet Union, and the UN sanctions against Iraq and Serbia, much emphasis has been placed on achieving international cooperation. The object is to deny the target country access to the supplies or markets of its principal trading partners and particularly when international organizations are involved to invoke the moral authority of the community of nations against the target. How-ever, the degree of cooperation realized has usually disappointed the lead country. Even in World Wars I and II, when the Allies ultimately achieved a high degree of cooperation, Germany was able to draw on supplies from Eastern Europe and adjacent neutral powers. The following statement, taken not from President Bill Clinton’s difficulties in maintaining strict sanctions against Iraq or from President George W. Bush’s problems in erecting a sanctions regime against Iran but from a commentary on World War I, describes the problem:
. . . all attempts in this direction [of a permanent inter-Allied organization] had been wrecked by the contradictory nature of the commercial interests of the Allied nations, which were only kept in touch with one another by means of intermittent conferences. . . . (Guichard 1930, 67)
That said, the extent of international cooperation in sanctions episodes increased sharply with the end of the Cold War, and the proportion of uni-lateral US sanctions declined sharply. This decline partly reflects a shift in the regional locus of many sanctions from the US backyard in Latin Amer-ica to the European backyard in AfrAmer-ica and in part to increased activity by the United Nations (see chapter 5). But—as the sanctions against Haiti, Sudan, and Iran show—the end of the Cold War guarantees neither that
58 ECONOMIC SANCTIONS RECONSIDERED
cooperation will be forthcoming nor that cooperative sanctions will work.
Although a complete economic blockade is seldom achieved, there are substantial differences from episode to episode in the degree of coopera-tion realized. We have used an index scaled from 1 to 4 to grade the ex-tent of cooperation:
1. No cooperation: A single sender country imposes sanctions and seeks (or gets) no cooperation; illustrated by the US campaign against Brazil to destabilize President João Goulart (Case 62-1) and by India’s campaign against Nepal for its pro-China tendencies (Case 89-1).
2. Minor cooperation: The sender country enlists verbal support and possi-bly token restraints from other countries; illustrated by the US sanc-tions imposed on the Soviet Union in part for its support of repressive measures in Poland (Case 81-3) and by US sanctions imposed against India for its nuclear tests (Case 98-1).
3. Modest cooperation: The sender country obtains meaningful restraints—
but limited in time and coverage—from some but not all the important trading partners of the target country; illustrated by the US sanctions against Castro’s Cuba in the early phases of that drawn-out episode and by US sanctions against Iran during the hostage crisis of the 1970s.
4. Significant cooperation: Important commercial nations make a major and coordinated effort to limit trade and/or finance, although the sanctions scope may still be limited (e.g., several UN arms embargoes), or signif-icant leakages may occur through neutral countries; illustrated by the two world wars, the early years of CoCom, and UN sanctions against Iraq and Serbia.
In addition to ad hoc cooperation, senders sometimes seek the coopera-tion or endorsement of an internacoopera-tional organizacoopera-tion. Tables 3A.1 through 3A.5 in chapter 3 record such endorsements, in instances when they oc-curred, by an abbreviation that indicates the cooperating organization. In addition, we created a variable indicating endorsement by an interna-tional organization of which both sender and target were members. These cases are indicated by marking the international organization’s name in bold in the chapter 3 tables. From the standpoint of the principal sender, enlisting an international organization probably lowers the “transactions costs” of persuading other countries to support a sanctions initiative.
Equally important is the greater aura of legitimacy conferred upon sanc-tions authorized by the United Nasanc-tions or (in an earlier era) the League of Nations and the consequent political isolation of the target country. How-ever, the humanitarian toll resulting from the near total embargo of Iraq has tarnished the UN “halo”; consequently it may take several years be-fore the United Nations will again authorize comprehensive sanctions (as contrasted with targeted measures).
ANALYZING THE UTILITY OF SANCTIONS 59 Moreover, we think that the many efforts and the inevitable failures in building watertight economic barriers have led to an exaggerated empha-sis on the mechanical role played by cooperation in determining the suc-cess or failure of a sanctions episode. Proponents of economic sanctions often engage in a wishful “if only” form of argument: “if only” the United States would stop all commerce with South Africa (in the apartheid era);
“if only” the Japanese would restrict their financial ties to China (in the af-termath of Tiananmen Square); “if only” the Europeans would deny Iran financial and trade links (in the context of Iran’s attempted acquisition of nuclear weapons).7 But the effort to create watertight barriers may be doomed from the beginning. And in many cases the symbolic moral au-thority of international cooperation exceeds the mechanical importance of stopping all trade and financial leakages.
Of course, from the sender country’s standpoint, it is axiomatic that more cooperation is better than less—whether the cooperation involves strong sanctions or moral condemnation. But international cooperation is seldom decisive, since other variables are also at play. A critical variable is the nature of the objective. The inspiring words of Robert Browning seem written for sender countries: “A man’s reach should exceed his grasp, or what’s a heaven for?” The pursuit of more ambitious objectives accompa-nied by much fanfare often goes hand-in-hand with efforts to enlist inter-national cooperation; yet the grasp of ambitious objectives may remain beyond the reach of sender countries, even when assisted by a large mea-sure of international cooperation.
International Assistance to the Target Country
The mirror image of international cooperation with the sender country is the support received by the target country from its neighbors and allies.
Target countries are seldom cut off from all alternative markets or financ-ing sources when sanctions are imposed; trade and financial channels usually remain open, even if at a sharply higher cost. For this reason, we do not count evasive and covert trade and finance as “assistance.” Such transactions are part and parcel of every episode. Rather we are con-cerned with overt rhetorical support accompanied by economic or mili-tary aid to the target country, in response to the imposition of sanctions.
The impact of sanctions on the target country can be reduced if it can rely on its friends to compensate for the burdens imposed by sanctions.
Further, the psychological rebuke is considerably weakened when third countries provide assistance to the target. Indeed, in several Cold War cases, target countries turned sanctions to their economic advantage, coax-ing opponents of the sender country to provide new or additional funds in
7. As an example of the “if only” argument in the Rhodesian context, see Brown-John (1975, 378).
60 ECONOMIC SANCTIONS RECONSIDERED
order to “make a stand” against the policy excesses of the rival power. The United States and its allies came to the rescue of Yugoslavia in the early postwar period when Josef Stalin threatened Tito. The Soviets similarly joined forces with Colonel Haile-Mariam Mengistu in war-torn Ethiopia to deflect US attempts to foster human rights and gain compensation for ex-propriated property (Case 77-8). In both cases, the amount of aid provided to the target more than offset the economic impact of the sanctions. In many episodes—such as the Soviet efforts against Albania, US efforts against Nicaragua and Iran, and UN sanctions against Serbia—assistance from a major foreign power provided welcome moral support to the tar-get. While we do not scale the degree of international assistance among our political variables in measuring the cost of sanctions to the target, we do take into account the estimated value of offsetting assistance. Among the political variables, however, we simply identify those cases where signifi-cant assistance was given to the target country.
Prior Relations Between Sender and Target
Sanctions are imposed against friends and foes alike. Forceful sanctions may be needed against belligerent countries to coerce them into yielding, especially since the stakes often involve national security or other major is-sues for both sender and target and because the target may be concerned about the reputational costs of conceding to a rival’s demands (Drezner 1999). On the other hand, a friendly country will often consider the im-portance of its overall relations with the sender country before responding to economic sanctions. In addition, a sender is likely to trade more with an ally, and therefore have more leverage, than with an adversary. Such con-siderations led South Korea and Taiwan to accede to mild US pressure and forgo construction of nuclear reprocessing plants in the mid-1970s (Cases 75-1 and 76-2). Likewise, the decertification process had a galvanizing ef-fect in turning the political tide against President Ernesto Samper in Colombia. With friends, subtle or symbolic sanctions may succeed.
To reflect the role of prior relations in determining the outcome of a sanctions episode, we have constructed an index for classifying the cases according to the state of political relations between the sender and target countries before the imposition of sanctions:
1. Antagonistic: The sender and target countries are in opposing camps; il-lustrated by most Cold War cases, US-Japan relations prior to World War II, Arab-Israeli relations, and US relations with Cuba, North Korea, Iran, and Libya for the past two decades or more.
2. Neutral: The sender country does not have strong ties to the target, but there is a workable relationship without antagonism; illustrated by im-mediate post–World War II relations between Spain and the United Kingdom despite centuries of dispute over Gibraltar (Case 54-3), US
re-ANALYZING THE UTILITY OF SANCTIONS 61 lations with Haiti prior to the 1987 sanctions (Case 87-2), and US rela-tions with Iraq in the late 1980s prior to the invasion of Kuwait.
3. Cordial: The sender and target countries are close friends and allies;
illustrated by ties between the Arab League and Egypt prior to the Camp David accords (Case 78-6), US relations with the United King-dom before the Suez crisis of 1956, Indian relations with Nepal before the 1989 dispute, and UK relations with Malta prior to base negotiations (Case 71-2).
Democracy versus Autocracy
To evaluate the target country’s political regime type we rely on the Polity IIId database created by Ted Gurr and colleagues and described in McLaughlin et al. (1998). The Polity project focuses on five dimensions of a political system’s authority: “(1) the influence relations between super-ordinate and subsuper-ordinate strata; (2) the degree of inequality between the strata; (3) the institutional relations among superordinates; (4) the com-petitiveness of recruitment to superordinate positions; and (5) the basis of political legitimacy, whether personal, substantive or procedural” (Gurr and Jaeger 1995, 470). This analytical framework provides the basis for the construction of indicators of regime type, regime coherence, and regime durability.
In terms of regime type, Gurr argues that no sharp dividing line sepa-rates democratic and autocratic regimes, but each dimension can be measured independently. Focusing on the institutional dimensions of democracy, the indicators of democracy and autocracy are derived from subjective coding of the institutionalized competitiveness of political par-ticipation, the regulation of political parpar-ticipation, the openness and com-petitiveness of executive recruitment, and the institutional constraints on the exercise of executive power. Because most regimes have mixed author-ity characteristics, Gurr and Jaeger (1995) established a single summary measure of the institutional characteristics of political regimes by subtract-ing a state’s autocracy score from its democracy score. The resultsubtract-ing index ranges from +10 for states that are purely democratic to –10 for completely autocratic states. We have adapted this variable for our regime type vari-able in this study, compressing the Gurr and Jaeger index to our own scale, running from 1 (autocracy) to 3 (democracy).
Economic Variables
Estimating the Economic Costs to Targets
Sanctions are designed to penalize the target country for its unwanted behavior. In theory, the target country will weigh the costs imposed by the
62 ECONOMIC SANCTIONS RECONSIDERED
sanctions against the benefits derived from continuing its policies—the higher the net cost, the more likely the target country will alter its policies.
The absolute cost exacted on a target country, however, is not the best measure of the potential impact: A cost of $100 million means more to Chile, for example, than to China. We have therefore related our estimated cost figures to the target country’s GNP. Appendix C explains in detail our methodology for estimating the cost to the target country.
Country Size and Trade Linkages
Quite apart from the magnitude of costs that the sender imposes on the target, the outcome of a sanctions episode may be influenced by the rela-tive size of the two countries and the trade links between them. The im-position of even minor sanctions carries the implicit threat of more dras-tic action. Whether that threat looms large or small depends very much on relative country sizes and trade flows. Hence, we include among our eco-nomic variables a ratio between sender-country and target-country GNP levels and figures on trade between target and sender expressed as a per-centage of the target country’s total trade.
Economic Health and Political Stability of the Target Country
The economic and political atmosphere in the target country also shapes the outcome of a sanctions episode. An analogy with rainmaking is ap-propriate. If storm clouds are overhead, rain may fall without anyone’s help. If moisture-laden clouds are in the sky, chemical seeding may bring forth rain. But if the skies are clear and dry, no amount of human
The economic and political atmosphere in the target country also shapes the outcome of a sanctions episode. An analogy with rainmaking is ap-propriate. If storm clouds are overhead, rain may fall without anyone’s help. If moisture-laden clouds are in the sky, chemical seeding may bring forth rain. But if the skies are clear and dry, no amount of human